Re: XMLLiteral

>In RDF Semantics, rdf:XMLLiteral is said to be a built-in datatype with
>special meaning.
>
>It is also said that datatypes are classes. Hence it is possible to define a
>subclass of a datatype.

True, but a uriref which denotes a subclass of a datatype class need 
not itself be a datatype uriref, and in fact usually will not be. 
Certainly, just asserting it to be a subclass does not thereby make 
it into a datatype.

>Now, if one defines a new datatype  as a subclass of rdf:XMLLiteral  :

That is a slightly misleading way to phrase things. Defining a 
subclass, and indicating that something is a datatype, are distinct 
assertions which are not simply related to one another.

>ex:MyXMLLiteral rdf:type rdfs:Datatype
>
>ex:MyXMLLiteral rdfs:subClassOf rdf:XMLLiteral
>
>Does ex:MyXMLLiteral play the same role as rdf:XMLLiteral ?

The second subClassOf assertion alone is sufficient to enable any 
triple using your property name to entail a similar triple using 
XMLLiteral:

ex:MyXMLLiteral rdfs:subClassOf rdf:XMLLiteral  .
aaa ex:MyXMLLiteral  bbb .

entail

aaa rdf:XMLLiteral bbb .

by rule RDFS 6, so in this sense your property can indeed play the 
same role as rdf:XMLLiteral (and there is no need for it to be a 
datatype in order to do this).

But if you mean more than this by 'play the same role', ie that there 
is a licence to use it directly as a datatype in a typed literal, 
then the answer to this is no.  RDF does not itself provide a 
mechanism for describing or defining new datatypes.  The first 
assertion given above,

ex:MyXMLLiteral rdf:type rdfs:Datatype  .

when suitably interpreted, means that your uriref 'ex:MyXMLLiteral' 
identifies a "recognized datatype", which (in operational terms) is 
an indication to an RDF engine that  information about that datatype 
should somehow made available from an external source by using the 
uriref.  So the only likely effect of this assertion at present might 
be to produce an error message of some kind to the effect that the 
datatype cannot be found.

Does this adequately respond to your query?

Pat Hayes

>In this case, the semantics should take this into account, for example here :
>
>''Since language tags play no role in the meaning of a typed literal, they can
>in practice be ignored, and any literal of the form "sss"@ttt^^ddd, where ddd
>is not rdf:XMLLiteral, treated as identical to the same literal without the
>language tag, "sss"@ddd.''
>
>Olivier Corby
>
>---
>Olivier Corby, Acacia project, INRIA Sophia Antipolis
>email : Olivier.Corby@sophia.inria.fr tel : +33 4 92 38 78 71
>http://www.inria.fr/acacia/personnel/corby


-- 
---------------------------------------------------------------------
IHMC					(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola              			(850)202 4440   fax
FL 32501           				(850)291 0667    cell
phayes@ai.uwf.edu	          http://www.coginst.uwf.edu/~phayes
s.pam@ai.uwf.edu   for spam

Received on Friday, 7 February 2003 14:20:12 UTC